Abstract:Data-driven process monitoring based on latent variable models are widely employed in industry. This article proposes a novel monitoring framework for latent variable models using hierarchical feature extraction, Bayesian inference, and weighting strategy. We first establish a deep structure to implement hierarchical latent variables extraction, the extracted features are used to construct diverse monitoring statistics. Then, we utilize Bayesian inference and proper weighting strategy to fuse various useful information. In line with the different characteristics of principal component analysis (PCA) and independent component analysis (ICA), we construct a deep PCA-ICA model for process monitoring according to the proposed framework. The deep PCA-ICA model performs hierarchical feature extraction, which can simultaneously extract deep Gaussian information and deep non-Gaussian information. The features extracted by different layers are then transformed to posterior probabilities through Bayesian inference. After that, different posterior probabilities are combined through appropriate weighting strategy to build new probabilistic statistic, which can give more synthetic monitoring results. Moreover, the Bayesian inference and weighting strategy are further used to integrate the advantages of different models by transforming various probabilistic statistics into an overall monitoring index, which can comprehensively indicate the process status. The Tennessee Eastman process is used to validate the superiority of the proposed model over the existing methods. Besides, the extracted features are further analyzed to show the effectiveness and benefits of the deep hierarchical feature extraction structure.

Autoregressive Dynamic Latent Variable Models for Process Monitoring

Dynamic Processes Modeling and Monitoring Based on a Novel Dynamic Latent Variable Model.

Statistical Monitoring of Dynamic Multivariate Processes Part 1. Modeling Autocorrelation and Cross-correlation

Quality-Relevant Process Monitoring with Concurrent Locality-Preserving Dynamic Latent Variable Method

Multimode Process Monitoring Based on Switching Autoregressive Dynamic Latent Variable Model

Autoregressive Double Latent Variables Probabilistic Model for Higher-Order Dynamic Process Monitoring

Variational Bayesian State Space Model for Dynamic Process Fault Detection

Robust Dynamic Latent Variable Model for Dynamic Process Monitoring with Missing Data in Cyber-Physical Systems

A New Method of Dynamic Latent-Variable Modeling for Process Monitoring

Investigation of Dynamic Multivariate Chemical Process Monitoring

Bayesian Robust Linear Dynamic System Approach for Dynamic Process Monitoring

Recursive Autoregressive Dynamic Latent Variable Model for Fault Detection of Dynamic Process with Missing Values

Switching Autoregressive Dynamic Latent Variable Model for Fault Detection in Multimode Processes

Dynamic Probabilistic Latent Variable Model for Process Data Modeling and Regression Application

Industrial Process Modeling and Fault Detection with Recurrent Kalman Variational Autoencoder

Kernel Dynamic Latent Variable Model for Process Monitoring with Application to Hot Strip Mill Process

Dynamic Process Monitoring Based on Variational Bayesian Canonical Variate Analysis

Enhanced dynamic latent variable analysis for dynamic process monitoring

Deep Learning of Latent Variable Models for Industrial Process Monitoring

Dynamic Latent Variable Modeling for Statistical Process Monitoring

New monitoring method for dynamic non-Gaussian process